hey folks,

 

when do u plan to update qemu to 4.0?

 

ty

+1

Better support for Q35 PCIe devices is enough for me to drool.

I would be satisfied if they add NUMAD first. I'm getting tired of having to restart unRAID so my VMs gets all the RAM from the right numa node. For me there are more performance benefits if the ram i'm getting is strictly from the cores i'm pinning to the vm's.

Edited July 26, 20196 yr by Nephilgrim

1 hour ago, Nephilgrim said:

I would be satisfied if they add NUMAD first. I'm getting tired of having to restart unRAID so my VMs gets all the RAM from the right numa node. For me there are more performance benefits if the ram i'm getting is strictly from the cores i'm pinning to the vm's.

QEMU 4.0 with better PCIe support is a matter of stability (and feasibility), which is a higher priority than performance, especially when your issue already has a workaround in place. 

With creative use of numatune + bash script, you can even allocate RAM to VM precisely across multiple nodes (e.g. if your VM tends to over-allocate on node 0 then have a dummy VM reserve an amount out of that node first using numatune (+ strict) then start the main VM then kill the dummy VM). I have been doing that for a while now and do not have to restart Unraid to get the right memory allocation.

Per-node process memory usage (in MBs) for PID 31136 (qemu-system-x86)
                           Node 0          Node 2           Total
                  --------------- --------------- ---------------
Heap                         0.02            0.04            0.05
Stack                        0.02            0.02            0.05
Private                  24412.36        24810.51        49222.87
----------------  --------------- --------------- ---------------
Total                    24412.40        24810.57        49222.96

If your gaming VM is only connected to 1 node then numatune + strict is sufficient and no workaround needed.

Edited July 26, 20196 yr by testdasi

QEMU 4.0 with better PCIe support is a matter of stability (and feasibility), which is a higher priority than performance, especially when your issue already has a workaround in place. 
With creative use of numatune + bash script, you can even allocate RAM to VM precisely across multiple nodes (e.g. if your VM tends to over-allocate on node 0 then have a dummy VM reserve an amount out of that node first using numatune (+ strict) then start the main VM then kill the dummy VM). I have been doing that for a while now and do not have to restart Unraid to get the right memory allocation.

Per-node process memory usage (in MBs) for PID 31136 (qemu-system-x86)                          Node 0          Node 2           Total                 --------------- --------------- ---------------Heap                         0.02            0.04            0.05Stack                        0.02            0.02            0.05Private                  24412.36        24810.51        49222.87----------------  --------------- --------------- ---------------Total                    24412.40        24810.57        49222.96

If your gaming VM is only connected to 1 node then numatune + strict is sufficient and no workaround needed.

That would be good if strict worked but the strict node 1 allocate my whole VM in node 0 and setting to preferred is the only way to, at least, allocate it where i want.



Enviado desde mi Mi MIX 2 mediante Tapatalk

QEMU updates arrive with major point releases.

 

I assume unraid 6.8 will have QEMU 4.0 or possibly 4.1 since it will be final soon.

Edited July 31, 20196 yr by Dazog

Can't wait for QEMU 4.0 or 4.1 ^_^ hopefully it improves performance with threadripper further.